ABSTRACT
This first part of the opening biblical quote is particularly appropriate for this chapter. From the previous chapters it is clear that it is not possible, in some mechanical manner, to obtain unbiased estimates of the age, period, and cohort parameters that generated the outcome values, even though these generating parameters are of central interest in the substantive fields that use age-period-cohort (APC) models.† For the individual age, period, and cohort coefficients there are an infinite number of estimates that fit the data equally well; therefore, we cannot use the likelihood of the estimated coefficients based on the data to determine the most likely generating parameters. Despite this limitation, I examine and extend a method that provides a partial view of the relationships between ages, periods, and cohorts to the outcome values. These estimates are unbiased estimates for the linear combination of the underlying data generating parameters.
